Author: Eran
Date: 15:24:37 03/14/02
Go up one level in this thread
On March 14, 2002 at 17:52:12, Peter Fendrich wrote: >There are a lot of test suites trying to compute ratings, like LCTII, BS2830, >BT2450, BT2630 and WM-test just to mention a few. >...but where are all the formulas???? > >I'm not a big believer of test suites to compute ratings but >I'm very curious :-) > >Peter There is no universal formula for all test sets. The purpose of any test set is to compare strength and playing style of several chess engines. For instance, running LCTII test set on Fritz7 and Shredder6 - 10 minutes of calculation per position equally for both of the chess engines. When it is finished, you can compare test set results of Fritz7 and Shredder6 - which chess engine solved most positions, etc. Test sets absolutely do not give you reliable Elo rating for various chess engines. Perhaps test sets are more accurate than chess engine tournaments but not perfect. Eran
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.